Research on the Initial Teaching Alphabet (i.t.a.) is viewed from two aspects--its limitations and how it can be improved. Thus far research has indicated that (1) transition from i.t.a. to traditional orthography is not the problem expected and (2) the children who are taught with i.t.a. read no better or worse than children taught with the conventional alphabet. Discussed are errors in research design related to (1) the use of control and experimental groups and materials used by each, the teachers, the investigator's attitude, sample size, and statistical analyses used; (2) the duration of the investigation, which in many studies is but one year; and (3) problems with criterion measures centering around such issues as final measures for evaluating i.t.a., the adequacy of reading measures, and correlated measures. Suggestions were made for improving research design which can be implemented as further studies replicate or build upon what has been done. Further research studies with i.t.a. could lead not only to its more effective evaluation but also to an examination of basic concepts relating to aspects of educational research. A bibliography is included. (DH)